Payment DataLake Reservoir System
- Problem Statement
-
- Build a single, complete, high quality store of all banking data.
- Business Goals
-
- For historic reasons the data was spread over a large number of payment and DDA systems. Every system “spoke a different language” in the way it represented data. In order to perform any kind of reporting, data science, processing (e.g. billing), operations, etc there was a need to consolidate the data.
- Implementation
-
- Publishing Framework
-
-
- Implementation of event, message and file publishing into the data lake using Apache Kafka
-
-
- Intelligence Engine
-
-
- Framework (ETL) for refining, transforming and segregating (for restricted countries) using Spark
-
-
-
- Spark Scala, Spark SQL, Spark Streaming, Spark DataFrames and Datasets
-
- Service Layer
-
-
- Data consumption via Kafka Topic
-
Enterprise Application Microservices to integrate various business services like (HR: Workday, Excelity, Benefits)
- Technologies included Java 8, JAX-RS, Spring Boot, Spring Core, ZUUL, Consul, Redis, OpenID, Connect, JWT, OAuth 2.0, SQL Server, ELK
- Amazon VPC, Docker
HIPAA compliant Connected Medical Devices with AWS IOT
- Telemetry data gets sent to AWS IoT Core using MQTT
- Device metadata gets stored in DynamoDB and device data is stored in S3.
- Batch ETL done using Amazon Glue and processed data is stored in S3